Establishment and evaluation of the sub-health diagnosis model based on decision tree

The paper introduces the idea of data mining on the basis of analyzing sub-health status quo and existing defects. An influencing-factor-model of sub-health state has been established with the help of the Microsoft decision trees technology, after evaluation, its accuracy rate reaches up to 89.81%. This model serves an important reference for the sub-health diagnosis, prevention and treatment, thus it has a wide range of practical value and significance.